The psychotic symptoms, delusions and hallucinations, are present in at least 30-40% of patients with Alzheimer's disease (AD). Psychotic symptoms in AD patients (AD+P) predict more rapid functional decline and premature institutionalization. Current treatments for AD+P are inadequate. We have hypothesized a polygenic model of AD+P. In preliminary tests of this hypothesis of this hypothesis, we have found that AD+P was significantly more frequent in patients with specific genotypes at the dopamine1 (D1) and D3 receptor loci (Sweet et a., 1998). Similarly, Holmes et al. (1998) reported an association in AD patients with variation in the serotonin/2A (5-HT2A) and 5-HT2C receptor genes. We propose to establish a cohort of 644 subjects, prospectively and longitudinally characterized with regard to psychosis phenotype, for examination of the genetic determinants of AD+P. Subjects with mild cognitive impairment, possible AD and probable AD will be evaluated at presentation to the Alzheimer's Disease Research Centers of the University of Pittsburgh and the University of Pennsylvania. Genetic material will be obtained. Neuropsychiatric assessments of psychotic symptoms will be conducted, with ratings on the CERAD Behavioral Rating Scale. Subjects without current or prior psychotic symptoms will be followed longitudinally with repeat assessments for psychotic symptoms every 6 months. Telephone assessments will be used for subjects unable to return to minimize incomplete data due to drop-outs. We project 20%-30% of the 644 subjects without psychosis at baseline will develop incident AD+P during the study interval. We hypothesize: 1) D1 receptor genotype will predict onset of AD+P; 2) D3 receptor genotype will predict onset of AD+P; 3) 5-HT2A receptor genotype will predict onset of visual hallucinations; 4) H-HT2C receptor genotype will predict onset of visual hallucinations. This study would be the first to prospectively evaluate the contribution of specific genes to predicting the onset of psychotic symptoms in any disorder. Replicated findings would provide a compelling rationale for family-based studies to address population stratification effects and for pursuit of the identified receptors as targets for drug development. Finally, establishing an AD cohort with prospectively determined psychosis phenotype will facilitate the research for novel risk genes as new genetic technologies become available.